reflect-paper

Warn

Audited by Snyk on Jul 7, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (medium risk: 0.65). The required workflow can ingest outsider-authored free text via delegated sibling skills that “fetch legal OA copies” (e.g., fetch-paper) and “study venue exemplars”/“find missing related work” (e.g., study-exemplars, find-papers, literature-review), which at runtime would read public web/published documents into the LLM context for the reflection passes.

Issues (1)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 7, 2026, 01:03 PM
Issues
1
Security Audit — snyk — reflect-paper